Venture Capital: Data Driven vs Traditional Strategies
The top venture capitalist of this age gets paid the big bucks because they claim they have a special eye for the billion dollar companies. They claim that they are able to identify early the best companies based on markets, team, product, trend or whatever it might be. This is that "gut" feeling. But that's the exact reason that traditional venture capitalists have failed to return good returns and that Venture capital as an asset class generally does not fare well against the market.
But what happen if we take a data-driven approach to venture capital?
So far I can think of a few firms that do this. Correlation VC, 500 Startups, Google ventures. Correlation VC actually takes historical data and invests based off of that. PAVS invests along the same lines. 500 startups involves both a quant and traditional approach in that they make smaller bets based off of gut instinct at the earliest stages.
Dave McClure of 500 startups had argued that due to low cost of startups these days, it's now possible to try to aim for larger number of base hit companies rather than that home run company. Steve Blank also argued that following a more data driven approach will statistically result in the success rate of raising capital and succeeding.
With the wealth of venture capital firms out there today, what strategy will represent the future of Venture Capital? Would love to hear some thoughts from VC industry people here.
Data driven works only if the future somewhat resembles the past. Real $ is made on black swan innovations and truly unique concepts. Just my 2 cents]
To be fair to the OP, it really depends on the industry.
Interesting post. I would echo what mb666 mentioned about the black swan events. The challenge and danger of using historical data to predict future performance is that we often forget to take into consideration the risk involved. Just as there can be upside to blackswan events, there can be large downsides, and more often than not, relying on historical data tends to hide those risks creating false sense of security. Very difficult to draw causative links between historical data and future performance. (Check out Nassim Taleb's writing on the topic)
VC all seems about the management of great teams. When working at such an early stage, having the right people in place often trumps the great idea. The ideas often change multiple times before the company becomes successful. Experimenting and learning quickly from you mistakes requires a great team, and I don't think historical data can help guide that decision.
I’ve spoken with both Correlation (Anu P.) and GV (Tyson C.) on their quantitative strategies. GV intrigues me most as it is what I try to use. According to related models I have seen, their take has more to do with systems theory and taking individual transaction data and developing a sensitivity analysis around it. Monte Carlo is used as well in some cases to discover mean scenario in a frequency distribution. Correlation’s take is proprietary according to those I’ve spoken with. But I have a hard time understanding their data sources for the predictive analytics that guides their decisions.
Sapiente aspernatur sit atque consequuntur praesentium. Et enim aliquid est deserunt molestiae nam mollitia explicabo. Nisi unde aut sapiente voluptas totam eius. Accusantium est quod voluptas eos. Porro veritatis consequatur rerum enim nesciunt unde.
Placeat animi vel quibusdam reiciendis voluptatibus. Similique odit fugit incidunt ut dolor dolore vitae. Pariatur ut accusantium sed sunt et soluta. Aliquam dolor et voluptatem voluptatem. Doloremque rerum nobis mollitia vel reprehenderit.
Non beatae labore enim sit facere. Qui voluptatem et minima dicta quia cumque rerum. Voluptatem iure minima dolorum impedit ut et.
Porro rem dolore quas rem inventore. Est hic labore alias et est voluptas saepe. In aut eum incidunt amet animi.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...